Scanning Electron Microscopes and Helium-ion Microscopes

SEMs and Secondary electrons For the investigation of

bulk materials

Electron probe, 1-2 nm (FE), scanning in a raster over a region of the specimen

The recorded signals are displayed in synchronism

Signals and information extracted • SE/BSE: topographic and material, magnetic field • SE: local potential • EBIC/CL: p-n junctions, lattice defects, quantitative measurement of semiconductor parameters • CL: concentration of trace elements • EDS/WDS: composition • BSE: crystallographic information (electron channelling patterns/Electron backscattering patterns)

Contrast: the electron energy, the specimen tilt or the collected angular range of emitted electrons, the detector configuration, energy selection

Secondary electrons

SEs: escape from a small depth 1-10 nm

Primary electrons

Excited by the PE- high resolution

Excited by BSE: larger diameter : • 0.1-1 um (10-20 keV) • 5- 50 nm (low energies)

Excited by BSE when striking parts of the specimen chamber (lower

polepiece)

SEs diffuse from the column to the chamber

A good detector: SE1 + SE2

Elastically Reflected

Low-loss electrons

Auger electrons

Everhart-Thornley(ET) Detector

A bias of ~ 10 kV

High gain and wide bandwidth up to 10 MHz system

Positively biased

4-eV SEs trajectories

Shadowing effect: only a fraction of SEs can be

detected

Topographic Contrast: SE

The dependence of the SE yield on the tilt angle of the local surface normal relative to the incident beam

An Everhart-Thornley detector on the right-hand side does not collect all SE, the signal is lower on the side opposite to the detector

Surface normal direct away from the detector: Darker

Electron diffusion with a diameter of the electron range ~0.05-10um; Near edge, more SE2 are excited by BSE leaving the side wall

Much lower scale when the distance from an edge is of the order of the exit depth ~ 0.5-20nm of the SEs

SE1 is ~ proportional to the mass-thickness the primary electron have to penetrate

sec0 Non-concentric Isodensities

10keV

30 keV

Apparent when the probe diameter is the order of the

exit depth tse ;

Particles < 1nm: increased SE signal

Material Contrast in SE mode

Local variation of the SE yield that cannot be attributed to differences in surface tilt

The yield increases monotonically with increasing atomic number

The increase of BSE is fast

The main contribution: SE2 excited by BSE

High energy > 5 keV: the SE image shows material contrast similar to that of the BSE

All SEs generated in the gold layer

Mangnetite, Fe3O4

Matrix 80% Si, 10% Al

SE BSE

Voltage Contrast Potentials generated by charging effects or biasing

Variations of electrostatic field between the specimen and the ET detector

Influence the SE trajectories and hence the SE intensity • Positively bias : darker

Voltage contrast on a MOS-FET by biasing the gate by -6V

L. Zhang, F. Gao and S. Huang, Journal of Microscopy, (2010).

Helium-ion Microscope

6th July, 2009

•Topography

•Material

Imaging

•Lithography

• Ion milling: graphene

•Beam induced chemistry

Fabrication

•Thin film measurement

•Particle identification

Spectroscopy

•Surface structure

•Work function

More?

Capability

Resolution < 0.75 nm at 45kV Field of View

Variable from 1 mm to 200nm Energy Spread

0.25-0.5 eV Beam current

1fA – 100 pA Sample size

50 mm in diameter x 25 mm thick

Detectors – Everhart-Thorley secondary

electron detector – Energy resolved backscattered

ion detector – Spectroscopic ally resolved

photon detector and manipulator

– Transmitted ion beam detector

•Insulating, Metal, Semiconductor

•Low-Z/high-Z

•Nanoparticles, thin films, bulk surface

Extended sample base

•Small Probe Size

•High secondary electron yield

•Short wavelength of He ions

•Small interaction volume

High spatial resolution

•Large depth of focus

•Surface sensitivity

•High Material contrast

•High topographic contrast

•Voltage contrast

•RBI spectra: sub-A film thickness

•RBI imaging: material/channelling contrast: defect analysis

More information

•Low beam damage (beam current: 0.1-10 pA)

•Smaller feature sizes

•Negligible contamination

•Low etching rate for soft/fragile materials

High precision Nanofabrication

Advantages of the Helium Ion Microscope

Semiconductor Applications • Imaging of defects • Mask inspection • Device metrology • Device failure analysis

SEMs

Spatial resolution

Surface charging

Surface contamination

Resolution Image fidelity

TEMs

Sample dimension

Small Sampling

FIBs

Spatial resolution: 10 nm (milling)

Contamination

Damage

The Ion Source – the Trimer

• Pyramid Tungsten tip • Three atoms on the top • Radius ~ 100 nm • Life span: days/weeks

Source formation: Thermo-field evaporation

•Diaphragm

•Smallest virtual source size: atomic •High brightness •Small energy spread: < 0.5 eV

25- 35 kV

Gas auto-ionization

acceleration

• Cryogenically cooled • Positively biased • Gas purification

•Absorption/desorption •W atom migration

Preserve the source shape

Destroy the source shape

• High ionization • Atomic tip

The ion source: better illumination

Beam

sources

Electron sources (100 kV) Ion sources

Tungsten LaB6 Field

Emission

ALIS He+ LMIS Ga+(21

keV)

Brightness

(A cm-2 sr-1)

~105 ~5×106 ~109 ~ 4×109 ~ 106

Energy

spread(eV)

3 1.5 0.3 0.2-0.5 50-100

Comparable to the FE electron gun : high-brightness, low energy spread, small probe size

Beam Current (fA - pA) Helium Gas

Pressure

High resolution The convolution of the probe size and the interaction volume

Secondary particles can also escape from different depths

Confined to the surface

No high energy backscattered electrons

Secondary electron by Ions: iSE iSE

PEE: Potential electron emission

Potential energy released from ion neutralization

Free electrons from solid

No lower energy threshold Significant: v < 107 cm/s

(He+, ~ 100 eV)

KEE: Kinetic electron emission

Kinetic energy of the ion

Free electrons from solid

For He+ (> 5 keV): PEE a few % of KEE

1st principal model Stopping Power

Scattering Cross-section Mean free paths

Monte Carlo Fit an accepted model

Generation of e: Berthe’s proposition

dS

dESE

1

SE yield: e/ion

Escaping of e: diffusion

d

zAzp

exp)(

probability The effective diffusion length

dzz

dz

dER

d

SE

0

exp5.01

A ~ 0.5

Total path length For electron: dE/ds = constant

Analytical for dE/dz From simulation

First principal Partial wave expansion Empirically, by fitting

Semi-empirical Quantitative: not good

d: Minimum resolvable feature,

Ion < electron

ESE = E

vion<<vBohr

Excitation: conduction electrons

Plamson Inner shell

2nd iSE imaging: topographical contrast

Image fidelity

Angle dependence

Clearer Profile

Lose sharpness Dark region

Small interaction volume

SE2/SE1 ratio Electron, E =25 keV: eSE2/eSE1 > 1 Poor resolution: SE2 emerges from a

region equal to the full width of the interaction volume

He+ ions: material/beam energy dependent For E > 50 keV, iSE2/iSE1 << electron case

EBackscattered ions < Eprimary ions

high resolution

Optimized He+ iSE: > 100keV

iSE yield +

iSE1/iSe2 +

Gun brightness +

Beam energy +

Appendix

The reduction of diffusion in LVSEM

SE: ION imaging Secondary ion imaging: compositional information with high spatial resolution

An understanding of the process of interaction between a primary ion beam and a solid target leading to the electron emission •Clean targets

• Contaminated surface: high yield • Low coverage: Change in work function – escape probability • High coverage: emission from the contaminates

• Two types of electron emission • potential emission

• Neutralization of the ion: Auger or resonance neutralization followed by Auger de-excitation – electron emission • Metal target: Emax = the first ionization energy of the ion – the work function of the target • For emission: Ei-2 >0 (Positive ions of inert gases, electronegative elements)

• kinetic emission • transfer of the kinetic energy of a bombarding ion to the electrons of a projectile-target system